Gaseous Complexes of Nickel Chloride with Aluminum Chloride and

Received October 13, 1981. The thermodynamic functions AH and AS of the reactions (A) NiC12(s) + LC13(g) = NiLCl,(g) and (B) NiC12(s) + L2Cb(g)...
0 downloads 0 Views 597KB Size
Inorg. Chem. 1982, 21, 2934-2938

2934

Contribution from the Institute of Inorganic Chemistry, University of Fribourg, PErolles, CH-1700 Fribourg, Switzerland

Gaseous Complexes of Nickel Chloride with Aluminum Chloride and Gallium Chloride F. P. EMMENEGGER,* P. FAVRE, and M. KLUCZKOWSKI Received October 13, 1981

+

+

The thermodynamic functions AH and AS of the reactions (A) NiC12(s) LC13(g) = NiLCl,(g) and (B) NiC12(s) L2Cb(g) = NiL2C18(g)(L = Al, Ga) have been determined by spectrophotometry of the gas phase from 300 to 840 OC and by quenching of the equilibrated gas phase and analysis of the condensates. From combination of the results with a previous study where L = In it is found that AS does not depend significantly on L but is approximately 70 J mol-' K-I for reaction A and approximately 43 J mol-' K-' for reaction B. For reaction A AH is most positive for L = Ga, making the Ga complexes the least stable ones of the series where L = AI, Ga, In. This stability sequence follows the dimerization energies of LCl,. The observed trend of the thermodynamic functions and the similarity of the spectra suggest that the structures of all NiLCls complexes on one hand and of all the NiL2Cls complexes on the other hand are similar.

The formation of gaseous complexes between di- and trivalent metal halides can be described by equilibrium 1. In + nLX3(g) = ML~X3n+2(g) M = alkaline-earth metal, transition metal L = Al, Ga, In, Fe, Sc X = C1-, Br-, In = 1, 2, 3 (4)

A12C16: log [Kdiss(bar)] = 6.655 - 5684/T- 1 6 0 7 0 0 / p

MXZ(s)

(1)

most cases the dominating equilibrium is that with n = 2. According to Dewing,' this is the case in the NiC12(s)/A1C13(g) system, where only NiA12C18(g)had been observed. But in the NiClz(s)/InC13(g) system, it has been foundZthat NiInC15(g) (n = l ) is the most important gaseous complex. It is therefore of interest to know whether the gaseous complexes of NiClZwith GaC13are intermediate between those with AlC13 and InC13, i.e., whether they are one of the rare examples where both ,equilibria, one with n = 1 and the other with n = 2, can be conveniently studied. Such examples are important in the understanding of the thermodynamics of the stepwise formation of gaseous complexe~.~ Evidence for the existence of gaseous complexes in the NiC12/GaC13system has been given by Schafer et a1.,6 but no quantitative studies have been performed so far. Dewing's investigation of the NiCl,(s)/AlCl,(g) system was part of a pioneering survey' and therefore not intended to be of ultimate accuracy. Furthermore, recent results of chemical transport experiments by Schafer3could not be reconciled with Dewing's values for the enthalpy and entropy of formation of NiAl,Cl,(g). It therefore seemed worthwhile to reinvestigate the formation of gaseous complexes between NiClZand AlC13. Experimental Section Chemicals. NiC12 was prepared by dehydrating the hydrated salt under vacuum and then subliming it in an evacuated quartz ampule. AIC1, (reagent grade, Fluka) was purified by sublimation under vacuum several times. GaCl, (prepared from 99.9% Ga) was by courtesy of Alusuisse and was used without further purification. Analysis. Al and Ga where analyzed by atomic absorption for small concentrations and by EDTA titration for large concentrations. The volume of ampules and spectrophotometric cells was calculated from the weight of water they could contain. Methods of Investigation. High-temperature spectroscopy and quenching experiments have been discussed p r e v i o u ~ l y . ~ ~ ~ (1) (2) (3) (4)

Computation. For the dissociation constants of L&,(g) vapor pressure of NiCI2, literature data have been used:

E. W.Dewing, Metall. Trans., 1, 2169 (1970). F.Dienstbach and F. P. Emmenegger, Inorg. Chem., 16,2957 (1977). H.Schifer and J. Nowitzki, 2.Anorg. Allg. Chem., 457, 13 (1979). A. Dell'Anna and F. P. Emmenegger, Helu. Chi". Acta, 58, 1145

and the (2)25

Ga2C16: log [&isS (bar)] = 7.072 - 4595/T

(3)'

NiC1,: log [ P N ~ (bar)] c ~ ~ = 10.136 - 12420/T

(4)2

The optical absorbance of the gas phase of our systems is given in eq 5 , m = monomer LCl,, d = dimer L2C16,c = complex, c,m = A =

nNiCl{ C N i C l z T

+%

nc.ml , m y

+%

nc,dl , d y

(5)

NiLCIS (complex with a monomer LCl,), c,d = NiL2Cls (complex with a dimer L2C16), V = volume of the spectrophotometric cell, I = optical path length of the cell (10 cm in all experiments), n = number of moles, and e = molar absorptivity (M-l cm-'). The first term in eq 5 is easy to calculate so long as there is solid NiC12 present (eq 4 and eNicI2 from ref 2 and 8). It is negligibly small below 600 OC. Depending upon temperature, AlC13 pressure, and wavelength, the second and third terms contribute appreciably to A . The number of moles of nickel in a cell is nNi

= nNiC12(s) + nNiCi2(g) + nNiLCls(g) + nNiL2Cls(g)

(6)

In all experiments only a small fraction of the LC13 is used to form gaseous complexes, and therefore only the first and second terms in eq 7 are important. Nevertheless, the third and fourth terms are taken into account, at least approximately (see below). Usually, the absorbance vs. temperature graph has a break when the solid phase, NiC12, disappears (e.g., Figure 3), but at the temperature of the break the vapor pressure of NiC12 is still given by eq 4. At this temperature the second term in eq 6 and the first term in eq 5 can be. calculated as mentioned above. For the determination of the molar absorptivity of a gaseous complex the second and third term of eq 5 have to be determined simultaneously. This is feasible if the experimental conditions can be chosen so that the absorbance of one complex strongly prevails at the break temperature. For the determination of the equilibrium constants of reaction 1, Kcsn= pcsn/Pg, and k:,d = Pc,d/P,', P, is calculated from the amount of LCl3 in the cell by using the law of ideal gases, &&2c16) and The calculation of P,,, or Pc,dfrom the absorbance is eq straightforward if only one of the two complexes absorbs light; otherwise approximations have to be used. The enthalpy and entropy of reaction 1 were calculated from the temperature dependence of the equilibrium constants (second law). AH and AS (and in some cases e) were fitted individually or simultaneously so that the optical absorbances calculated with these parameters and the constants of the cell gave a best fit to the observed ab~orbances.~The error limit given is twice the standard deviation;

7.

(\ 1 979 -- ' - I .

(5) F.Dienstbach and F. P. Emmenegger, J. Inorg. Nucl. Chem., 40,1299 (1978). (6) H.Schifer, M.Binnewies, W. Domke, and J. Karbinski, 2.Anorg. Allg. Chem., 403, 116 (1974).

0020-1669/82/1321-2934$01.25/0

(7) W.Fischer and 0. Jiibermann, 2.Anorg. Allg. Chem., 227,227(1936). (8) C.W.DeKock and D. M.Gruen, J . Chem. Phys., 44,4387 (1966). (9) C.Daul and J. J. Goel,J . Chem. Soc., Faraday Trans. I , 73, 985 (1977).

0 1982 American Chemical Society

Inorganic Chemistry, Vol. 21, No. 8, 1982 2935

Gaseous Complexes of NiC12 with AlCI, and GaC13

,o'

Table I. Samples for the Visible Spectroscopy of NiAlCl,(g) and NiAl$l,(g) amt, mmol samule V. cms 1 2 3 4

5 6

29.05 30.23 30.23 30.15 28.12 27.87

NiC1, 9.95 X 15.9 x 13.5 x 16.3 x 14.3 x 6.79 x

AlC1,

T , "C

SiC1,

2.61 4.33 1.43 2.94 0.315 0.304

460-650 427-506 461-557 405-550 300-838 402-850

no no yes yes yes yes

2.06pm-'

1.7

3.95 550'C 1.L 12

LSO'C

13

1L

18

8'

350°C 15

16

l&T

*

Figure 2. In (PNjAlzAII/PAIIC16) =f( l/T). Sample numbers refer to Table I.

of sample 5 at 3.95 pm-' was used to calculate K(8) at temperatures from 300 to 490 O C . (It should be mentioned that NiC12(s) + A12C16(g) = NiA12Cls(g) (8)

Figure 1. UV-visible spectra of NiC12(g)(UV from ref 8), NiAlCl,(g), and Ni&C18(g) (from ref IO).

systematic errors are not considered. Results NiC12/AIC13. Dewing' investigated equilibrium 1a by enNiCl,(s) 2AlC13(g) = NiAl+&(g) (14

+

trainment measurements in the range P = 0.3-2 bar and T = 400-600 OC. The UV-visible spectrum of NiA12Clg(g)at 750 K has been published by Papatheodorou.IO When we heated NiC12 and AlC13, the quartz windows of our optical cells were severely attacked at temperatures above -700 K. With use of a hint from Schafer," this quartz attack could be completely suppressed by adding about 100 mg of SiC14 to a 30-cm3 optical cell (no influence of SiC14on the spectra of the gaseous NiCl2/A1Cl3 complexes could be'detected). With samples containing a rather large amount of A1Cl3 (example no. 4 of Table I), we were able to reproduce Papatheodorou's spectrum of NiA12C18(g),but with samples containing less AlC13 (examples no. 5 and 6, Table I), the spectrum corresponded to that of Ni111Cl~(g).~We therefore concluded that NiA12Clg(g)was the dominant gaseous complex at low temperatures and high aluminum chloride pressures while NiAlC15(g)would prevail at elevated temperatures and smaller aluminum chloride pressures. The spectra of the gaseous nickel-containing species in the NiCl2/A1Cl3 system are shown in Figure 1. Formation of NiA12C18(g). The absorbances at 1.7 and 3.95 pm-l are characteristic of NiA12C18(g) (Figure 1). In the temperature range of 450-550 O C , K(8) was calculated from the absorbance of sample 4 at 1.7 pm-I, while the absorbance ~~~

~

(IO) G. N.Papatheodorou, J. Phys. Chem., 77, 472 (1973). (1 1) H.Schifer, Z.Anorg. Allg. Chem., 445, 129 (1978).

K( la) = K(8)/Kdiss(A12C16).) The temperature dependence of K(8) was used to calculate preliminary values of AH and AS of reaction 8. They were needed to estimate how much equilibrium 8 interferes in the evaluation of samples dominated by equilibrium 9. In samples 1-4, at low temperatures, the absorbance at 2.06 pm-' is mostly due to NiA12C18(g), but the absorbance of NiA1C18(g) is not negligible. When AH and AS of reaction 9 were known (see below), the absorbance of NiAICIS(g) in samples 1-4 could be calculated, and if it was less than 10% of the total absorbance, the data were used to calculate K(8). All the In K(8) from evaluation at 1.7, 2.06, and 3.95 pm-' are shown in Figure 2. Considering the range of temperature (350-550 "C) and pressure (1.3-4.2 bar of A12C16(g)),the linearity of the In K vs. 1/ T plot is satisfactory. From a linear least-squares fit it follows that at 450 OC AH(8) = 62.8 f 2.5 kJ mol-' and AS(8) = 38.1 f 3.8 J mol-' K-l. In order to verify the result of the optical investigation by an independent method, we performed three quenching experiments at 400 OC. They yielded K(8) = (1.6 f 0.4) X while with AH(8) and AS(8) from optical measurements, K(8) = 1.3 X is obtained. Considering the large error generally associated with quenching experiments, the agreement is satisfactory. Formation of NiAICIS(g). The equilibrium constants of reactions 9 and 9a are related to each other: K(9) = NiC12(s) AlCl,(g) = NiAICIS(g) (9)

+ NiC12(s) + 1/2A12C16(g)= NiAICIS(g)

(9a)

K(9a) [ 1/Kdis(M2cb)1/2].For convenience we will evaluate our measurements according to equilibrium 9. Sample 5 contains only a little aluminum chloride, and its absorbance at 2.1 pm-l is therefore almost entirely due to NiC12(g) and NiAICls(g) and not to NiA12Cls(g). The break in the absorbance vs. temperature graph (Figure 3) is due to the disappearance of the solid-phase NiC12, and the decrease of the absorbance above this temperature is caused by reaction 10 (at 2.1 pm-': eNiCIz < 6NiAIC15,see Figure 1). At 760 OC, NiAICls(g) = NiC12(g) + AlC13(g) (10) the temperature of the break, the composition of the gas phase iS nNjCIz = 4.24 x lo", nNhICI, = 9.98 x lo", and nNjAI2Cl8 0.051 X lo", from which t N ~ c=l S319 M-' cm-' at 2.1 pm-'. With this value for tNiAIC1,, K(9) is determined from the increasing (550-750 "C) and K(10) from the decreasing

2936 Inorganic Chemistry, Vol. 21, No. 8, 1982

Emmenegger, Favre, and Kluczkowski

Table 11. Enthalpy (kJ mol-') and Entropy (J mol-' K-') of Formation of NiLCl,(g) and NiL2C18(g)at 298 K

+ L a 3 @ )= NiLCl,(g) C, = -6 J mol-' K-' a (B) NiCl,(s) + L,Cl,@) = NiL,Cl,(g) C, = -4 J mol-' K-' (C) CoCI,(s) + L2C16(g)= CoL,Cl,(g) (for Comparison) (A) NiCl,(s)

a

L

98.8 ?I 2.5 64.5 f 2.gC 46.4d

A B C

Estimated value. ences 2, 18, and 23.

72.1 ? 2.5 41.6 2 4.2c 46.2d

Reference 2.

108.8 f 2.5 IO f 8 45.6e

Reference 1: Ah'= 51.9, A S = 37.7.

1-

I-//+ -

500

93.7 t 8.4 77.4 f 8.4 49.5f

67.1 2 8.4 45.2 f 8.4 43.55

Reference 1, 4, and 24. e Reference 12 and 22.

Refer-

Table 111. Samples for the Visible Spectroscopy of NiGaCl,(g)

A 2.06p1K'

0

67.4 t 2.1 42a 40.2e

a '

_ _ - -_ -

600

700

800

'C .c'

Figure 3. Temperature dependence of the absorbance due to the formation and decomposition of gaseous NiC1,/AlC13 complexes: (0) sample 5 ; (0)sample 6 ; (-) absorbance calculated with thermodynamic values of Table I1 and EN^^^^ = 3 19 M-I cm-', t ~ i =~ 1117~ M-I cm-', t N ~ 1 2 = ~ !80 8 M-' cm-' at 2.1 Fm-I; (-.-) absorbance of sample 6 neglecting formation of NiAl2Cl8(g);(- - -) absorbance of NiCl,(g) in equilibrium with NiCl,(s).

amt. mol GaC1,

sample

V , cm3

NiC1,

1 2 3 4

34.99 33.70 33.17 38.98

1.47 x 10-5 excess 4.45 x 10-6 1.39 x 10-5

1.29 x 1.10 x 1.10 x 1.41 x

10-3 10-3 10-3 10-3

Table IV.' Thermodynamics of 1/2Ni2C14(g)+ '/2L,C16Q) = NiLC1,Cg) AH"298,kJ mol-' J mol-' K-'

A1

Ga

In b

-13.0 -0.8

-18.8 -6.7

-15.5 -7.5

a Thermodynamic values used to convert values of Table 11, eq in kJ mol-', ASz9, in J mol-' A, into values of Table N (AH,,, K-'1: NiC12(s) NiCl,(g), 247.2, 216.3;, 2NiCl,(g) -+ Ni,Cl,(g) -144.8, -133.9;'? 2AlC1,Cg) -+ A1,C16(g) -126.5, -153.1;" 2GaCl,(g) .+ Ga,Cl,t) -94.6, -150.6;7*1721nC13(g) In,Cl,(g) -131.4, -152.3.'O Reference 2. -+

-f

(760-840 "C) absorbance. K(9) and K(10) are related to each other by AG(10) = AG,,bl(NiC1l)- AG(9). Under the experimental conditions of sample 5, equilibrium 11 can be neglected. In sample 6 the ratio of aluminium NiA12Cls(g) = NiAICIS(g) A1C13(g) (1 1) chloride to nickel chloride is larger than in sample 5, and therefore equilibrium 11 has to be considered (AG( 11) = AG(9) + AGdiss(AI2Cl6)- AG(8)). The joint evaluation of samples 5 and 6 yields M ( 9 ) = 94.8 f 2.5 kJ mol-' and AS(9) = 65.0 f 2.5 J mol-' K-' at T,, = 940 K. If we combine AG(8), AG(9), %iAIClr, and ~ i A l z c lwith s literature values for AGsubl(NiC12) and AGdiss(A12Cb) to calculate the absorbance vs. temperature in our samples 5 and 6, the agreement between calculated and measured absorbance is quite satisfactory (Figure 3). NiC12/GaC13. Formation of NiGa2C18(g). As expected, the stability of NiGa2Cls is OW,'^,'^ but it was not anticipated that its spectrum would be masked by the spectrum of NiGaCl,. As this was the case, it was not possible to determine the equilibrium constant of eq 12 by optical spectroscopy but only NiC12(s) + Ga2C16(g) = NiGa2Cls(g) (12) by quenching experiments, which are not very accurate. At 400 "C we obtained K(12) = 5.24 X With an estimated entropy ~ a l u e ' ~ofJ aS(12) ~ = 46 J mol-' K-', we obtained AH( 12) = 70 kJ mol-'. Formation of NiCaCl,(g). Equilibrium 13 could be invesNiC12(s) + GaCl,(g) = NiGaC15(g) (13) tigated by optical spectroscopy. The samples are described

+

(12) F. P.Emmenegger, Inorg. Chem., 16, 343 (1977). (13) J. W. Hastie, "High Temperature Vapors", Academic Press, New York, 1975, pp 126-147. (14) H. Schlfer, Angew. Chem., In?.Ed. Engl., 15, 713 (1976).

A

i

,11p.' NI CI,/

P

GaCI,

1.5

7

/"

P

1 '

650

700

750

BOO

'C

Figure 4. Temperature dependence of the absorbance due to the formation and decomposition of NiC12/GaC13complexes: (0)sample 1, (0) sample 2, (A)sample 3, ( 0 ) sample 4 (sample numbers refer to Table 111); (-) absorbance calculated with the thermodynamic = 2 15 M-I cm-', t ~ i c = l ~117 M-' cm-'. values of Table I1 and %ia~lJ

in Table 111. The spectrum of NiGaCl,(g) is similar to those of NiAlCl,(g) and NiInCl,(g): with a peak at 2.1 1 pm-'. The molar absorptivity, t2,11rm-~ = 215 M-' cm-', was calculated from the (extrapolated) maximum of absorbance in the A vs. T graph (Figure 4) of samples 1 and 4. In deriving K( 13) from absorbance measurements at 2.11 pm-', the formation of NiGa2Cls was taken into account (although it was unimportant) by using the approximate AH and AS values mentioned above and an estimated molar absorptivity of 80 M-' cm-' in analogy to the molar absorbtivity of NiA12C18(g). We find AH(13) = 104.4 f 2.4 kJ mol-' and AS(13) = 60.2 f 2.1 J mol-' K-' at Tay= 1020 K.

Inorganic Chemistry, Vol. 21, No. 8, 1982 2937

Gaseous Complexes of NiC12 with AlCl, and GaCl, Table V. Stepwise Formation of Gaseous Complexes

L = Ga

L=Al - u " z 9 8 ,

reaction

kJ mol-'

NiCl,(g) + LC13(g) = NiLCl,(g) NiLCl,(g) + LC13(g) = NiL,Cl,(g)

148.5 160.2

average

154.4 * 5.9

-As02981

J mol" K" 144.3 183.3

- ~ " 2 9 , ,

kJ mol"

L = In" -Aso298,

- u 0 2 9 , ,

J mol-' K-'

138.5 129.7

kJ mol"

149.0 176.1

J mol-' K"

156.9 144.3

134.1 f 4.4

150.2 166.5

150.6 f 6.3

" Reference 2. Discussion NiAlC15(g). NiA1C15(g) has been observed by Binnewie~'~ in the mass spectrometer. If we use his published ion intensities to calculate K(9), we obtain K(9)893~= 8.5 X lo-,, while our measurements yield K(9)893~= 9.9 X lo-,. Considering the limited accuracy of the two methods of investigation, the agreement is excellent but perhaps somewhat fortuitous. NiA12Cls. For equilibrium 8 (B in Table 11), our thermodynamic data are rather different from those of ref 1; e.g., at 773 and 873 K our K ( 8 ) is more than five times smaller than that of ref 1. Schafer, transported NiC12(s) with A12C16(g) from 736 to 676 K and observed a transport rate of (0.36 f 0.07) X mol h-l. The predicted transport rate computed with the thermodynamic data of ref 1 is around five times larger, (2 f 0.3) X 10-5 mol h-', while with our thermodynamic data it is (0.39 f 0.06) X mol h-', in good agreement with the experimental value. The contribution of equilibrium 9 to this chemical transport is less than lo%, and therefore the agreement between calculated and observed transport rate mainly concerns our data of equilibrium 8. Thermodynamics and Structures. For an interpretation of the thermodynamics of the formation of gaseous metal halide complexes, various models have been proposed, and our results shall be checked against them. (a) The entropy and enthalpy of reaction 14 are approxi-

'/2(NiC12)2(g) + '/2(Lc13)2(8) = NiLC15(g)

(14)

mately ~ e r o . ' ~ ,A' ~small negative enthalpy value could be rationalized by the reduction of the cation-cation repulsion energy.', To calculate the enthalpy and entropy of reaction 14, one relies on the data for the evaporation and dimerization of nickel chloride, which are both known only to ca. f 8 kJ mol-' and 8 J mol-' K-I, re~pectively.'~.'' Considering this source of uncertainty, our results agree reasonably well with the condition set by equilibrium 14 (see Table IV). (b) Because the number and type of bonds formed in reactions 15 and 16 are similar, the enthalpies of reactions 15

Table VI. Stability of NiL,Cl,(g) and NiLCl,(g) at 900 K K(17)=Pc,d/P, K(1a) =Pc,d/Pm' K(~)=P,,,/P, udim(LCl,)"

A1

Ga

In

2.2 X lo-' 1.8 X lo3 7.8 x 10-3 -127

1.1 X lo-' 1.2 X 10' 1.2 x 10-3 -95

0.6 X lo-' 6.4 X 10' 9.1 x 10-3 -131

kJ mol-'; see Table IV. bar

f

Figure 5. Total pressure of nickel complexes in an ampule containing 1 mol of LCI3/22.4 L and excess NiCI,: 1, NiAICls + NiAl,C18; 2, NiInCIS NiIn2C18; 3, NiGaCIS NiGa2CI8; 4, NiC1, for com-

+

+

parison.

(16) = -170 J mol-' K-'. So far, most examples used to illustrate this scheme of stepwise formation of gaseous complexes have been complexes with indium chloride, but the complexes of nickel chloride with aluminum and gallium chloride fit as well (Table V). It should be mentioned that according to this model, the relative thermodynamic values of reactions 15 and 16 are not altered if MLX5(g) contains a fourfold coordinated M (trigonal pyramid) and ML2X,(g) a sixfold coordinated M (octahedron). (c) In a recent review, Schafer emphasizes the importance of changing (or maintaining) the coordination number of M in the course of the formation of gaseous metal halide complexes.2' In reaction 17, the change of coordination M suffers MC12(s) + L2c16(g) = ML2C18(g)

(17)

in going from the solid to the gaseous phase does not depend on L. Therefore, the thermodynamic functions of reaction 17 should, for a given M, be rather independent of L. The complexes of CoClz have so far been the only ones where the stability with all three group 3 chlorides (AlCl,, GaCl,, and InCl,) have been known, and these results are included in table I1 for comparison. Considering the limited accuracy of our data (estimated AS) for the formation of NiGa2C18(g) and and 16 should be ~ i m i l a r . ~ The . ' ~ entropies of these reactions will depend little on the masses of L and M, and they have been estimated5J8to be AS(15) = -150 J mol-' K-' and AS(15) M. Binnewies, Z . Anorg. Allg. Chem., 437, 25 (1977). (16) 0. Kubaschewski, L. L. Evans, and C . B. Alcock, "Metallurgical Thermochemistry", Pergamon Press, Oxford, 1967. (17) H. Schtifer and M. Binnewies, Z.Anorg. ANg. Chem., 410,251 (1974). (18) F. Dienstbach and F. P. Emmenegger, Helu. Chim. Acta, 60, 166 (1977).

(19) I. Barin and 0. Knocke, 'Thermochemical Properties of Inorganic Substances", Springer-Verlag, Berlin/Heidelberg, 1973. (20) 0. N. Komshilova, G. J. Novikov, and D. G. Polyachenok, Russ. J. Phys. Chem. (Engl. Trans.), 43, 1680 (1969). (21) H. Schafer, Adu. Inorg. Chem. Radiochem., in press. (22) A. Anundskas, E. A. Mahgoub, and H. A. 0ye, Acra Chem. Scand., Ser. A , 30A, 193 (1976). (23) G. H. Kucera and G. N. Papatheororou, J . Phys. Chem., 83, 3213 (1979). (24) G. N. Papatheororou, Z . Anorg. Allg. Chem., 411, 153 (1975). P. J. Thistlethwaite and S. Ciach, Inorg. Chem., 14, 1430 (1975). (25) H. A. Oye and D. M. Gruen, J . Am. Chem. SOC.,91, 2229 (1969).

Inorg. Chem. 1982. 21. 2938-2945

2938

NiIn2C18(g),the hypothesis that equilibrium 17 does not depend on L seems to be supported by the nickel complexes (Table 11). If we calculate K( 17) at, e.g., 900 K using the thermodynamic functions of Table 11, we find that the values differ by less than 1 order of magnitude (Table VI) and that KNiA12Cls > KNiGagb> KNihplclS. This, however, does not imply that the partial pressure of the gaseous nickel complexes in an ampule containing solid nickel chloride and identical amounts of LC13 is proportional to these equilibrium constants. To illustrate the relative stability of NiL2C18,it is better to compare the reactions NiCl,(s) + 2LC13(g) = NiL2C18(g),which show that the Stability of NiL,C18(g) is roughly proportional to the dimerization energy of LC13(g). The same trend is observed for the formation of NiLC15(g). This is expected if the approximations discussed in the context of eq 15 and 16 are valid. The relative amounts of nickel carried into the gas phase under "reasonable" experimental condition are shown in Figure 5 . Schafer2' and Dewing' proposed that the thermodynamics of reaction 17 should be similar ( A H = 50 f 8 kJ mol-', AS = 46 f 13 J mol -I K-I) for all M which have octahedral coordination in the solid MCl, as well as in the gaseous complex. While the enthalpies of formation of the cobalt complexes fit into this scheme, the enthalpies of the nickel complexes do not-they are too positive (Table 11). Within Schafer and Dewing's concept this would be understandable if the coordination number of the nickel was smaller in the gas than in the solid. This, however, is contradicted by the UV-visible spectrum, which is essentially that of a NiCl, chromophore1°

with only a small-if any-contribution from a tetrahedral NiCl, center. Although we have no proposition for reconciling the low stability of the NiL2C18(g) complexes with the structure derived from their optical spectrum, we consider it worth mentioning that NiClz has by far the highest melting point (1000 "C) of all the isostructural chlorides (CdCl,-type) for which equilibrium 17 has been studied (e.g., MnCl,, 650 "C; CoCl,, 740 "C). The structural stability which is responsible for the high melting point of NiC12 evidently also is against its "evaporation" by equilibrium 17. The results of the present investigation of the stability of NiA1C15(g), NiA12C18(g),NiGaC15(g), and NiGa2C18(g)are largely consistent with the ideas about the correlation between stability and structure of gaseous complexes. This agreement with the general aspects of a larger experimental body of gaseous complexes is considered as an additional element of support for our results. Acknowledgment. The authors thank Dr. C. Daul and J. Kuenlin for computer assistance. The project has been supported by the Swiss National Science Foundation, Grant No. 2.654-0.80. Note Added in Proof. While this paper was in press, a paper dealing with the stability Of NiAl,Cl&) and NiAl,Cl,,(g) has been published by W. Lenhard and H. Schafer, Z . Anorg. Allg. Chem., 482, 167 (1981).

Registry No. NiC12, 77 18-54-9; AlCI,, 7446-70-0; GaCl,, 13450-90-3; NiAlCI,, 81315-93-7; NiGaCI,, 66143-12-2; NiAl,C18, 40556-06-7; NiGa,C18, 66594-41-0; A12C1,, 13845-12-0; Ga2C1,, 15654-66-7.

Contribution from Rocketdyne, a Division of Rockwell International Corporation, Canoga Park, California 91304

Fluorine Perchlorate. Vibrational Spectra, Force Field, and Thermodynamic Properties KARL 0. CHRISTE* and E. C. CURTIS

Received December IS, 1981 Infrared spectra of gaseous, solid, and matrix-isolated C10,OF and Raman spectra of liquid C1030F are reported. All 12 fundamental vibrations expected for the covalent perchlorate structure

HC' 00

'

of symmetry C,were observed and assigned. A modified valence force field was computed for C10,OF by using the observed 35C1-37C1isotopic shifts, symmetry relations between the A' and the A" block, and the off-diagonal symmetry force constants of the closely related FC103 molecule as constraints. Previous assignments for C1030C1, C1030Br, C1O30CF3,C1207, and C1,07 are revised. The I9F N M R spectrum of C10,OF was recorded, and thermodynamic properties were computed in the range 0-2000 K.

Introduction Fluorine perchlorate (or perchloryl hypofluorite) was probably first prepared' in 1929 by Fichter and Brunner by the fluorination of dilute HClO, with F2 but was incorrectly identified. The first positive identification of C1030F was reportedZin 1947 by Rohrback and Cady, who obtained the compound from the reaction of F2 with concentrated perchloric acid. They reported that C1030F consistently exploded when frozen. (1) (2)

Fichter, F.; Brunner, E. Helo. Chim. Acta 1929, 12, 305. Rohrback, G. H.; Cady, G. H. J. Am. Chem. SOC.1947, 69, 677. 0020-1669/82/ 13 2 1-293 8$0 1.25 /O

In view of its explosive nature, it is not surprising that very few papers dealing with C1030F have been published since then. In 1962, Agahigian and coworkers reported3 the I9F N M R spectrum of C1030F ih CFC13 and four infrared absorptions of the gas. The same four infrared bands have also been observed in a studf at United Technology Corp. in which the heat of hydrolysis was measured for C1030F. Macheteau and Gillardeau studied5 the thermal decomposition of ClO,OF (3) Agahigian, H.; Gray, A. P.; Vickers, G . 0. Can. J . Chem. 1962.40, 157. (4) Brazeale, J. D.; et al. "Thermochemistry of Oxygen-Fluorine Bonding", Report UTC 2002-FR, AD No. 402889; United Technology Corp: Sunnyvale, CA, March 1963. 0 1982 American Chemical Society